Usefulness of Q-Wave Area for Threshold-Based Stratification of Global Left Ventricular Myocardial Infarct Size

Left ventricular (LV) infarct size affects prognosis after acute myocardial infarction (AM!). Delayed enhancement cardiac magnetic resonance (DE-CMR) provides accurate infarct quantification but is unavailable or contraindicated in many patients. This study tested whether simple electrocardiography (ECG) parameters can stratify LV infarct size. One hundred fifty-two patients with AM! underwent DE-CMR and serial 12-lead ECG. Electrocardiograms were quantitatively analyzed for multiple aspects of Q-wave morphology, including duration, amplitude, and geometric area (QWAr) summed across all leads except aVR. Patients with pathologic Q waves had larger infarcts measured by DE-CMR or enzymes (both p <0.001), even after controlling for infarct distribution by CMR or x-ray angiography. Comparison between early (4 +/- 0.4 days after AMI) and follow-up (29 +/- 6 days) ECG demonstrated temporal reductions in Q-wave amplitude (1.8 +/- 1.4 vs 1.6 +/- 1.6 mV; p = 0.03) but not QWAr (41 +/- 38 vs 39 +/- 43 mV.ms; p = 0.29). At both times, QWAr augmented stepwise with DE-CMR quantified infarct size (p <0.001). QWAr increased markedly at 10% LV infarct threshold, with differences more than threefold on early ECG (59 +/- 39 vs 18 +/- 20 mV.ms; p <0.001) and nearly fivefold (59 +/- 46 vs 13 +/- 16 mV.ms; p <0.001) on follow-up. Diagnostic performance compared with a 10% infarction cutoff was good on early (area under the curve = 0.84) and follow-up (area under the curve = 0.87) ECG. Optimization of sensitivity (95% to 98%) enabled QWAr to exclude affected patients with 90% to 94% negative predictive value at each time point. In conclusion, LV infarct size is accompanied by stepwise increments in Q-wave morphology, with QWAr increased three-to fivefold at a threshold of 10% LV infarction. Stratification based on QWAr provides excellent negative predictive value for exclusion of large (>= 10%) LV infarct burden. (C) 2013 Elsevier Inc. All rights reserved.